International Journal of Computer Applications (0975 – 8887) Volume 124 – No.17, August 2015 20 A Real Time Visible Watermarking Technique using Dual Tree Complex Wavelet Transformation Prodipta Bhowmik Assistant Professor, Department of IT Techno India, Salt Lake Kolkata-700091, India Tanmay Bhattacharya, PhD Associate Professor, Department of IT Techno India, Salt Lake Kolkata-700091, India ABSTRACT The key idea of this paper is to propose a real-time, self- authenticating, watermarking technique. Here the watermark has been constructed from the image itself using Toeplitz matrix and the embedding region is selected by the user at real-time. Embedding region should be selected from Region of Non Interest (RONI) of the image. Dual Tree Complex Wavelet transformations (DT-CWT) based embedding technique has been applied for the embedding purpose. The key is shared or transferred to the receiver through a secret channel. At the receiving end, the extracted watermark is matched with the matrix generated from the bit-sequence (key) which ensures the originality of the image. Keywords Digital watermarking, Toeplitz matrix, DT-CWT, Image authentication. 1. INTRODUCTION In today’s world, the use of digital multimedia content has increased a lot. Images can be processed, used and shared very easily. It can be transmitted also. Hence there are risk of unauthorized access and modifications of digital content. So authentication procedure needs to be applied to secure digital contents. Watermarking is a technique to hide digital information into other digital contents. It verifies the authenticity or integrity of the carrier signal. A digital watermark should be robust with respect to various attacks if the embedded information may be detected reliably even if degraded by various noises. A watermarked signal is called imperceptible if it is equivalent to the original one. A self authentication technique is basically used to identify tamper detection. The digital watermarking can be categorized into two types mainly: Spatial domain and Frequency domain. Least significant Bit (LSB) [1] is the most common spatial domain technique. Intensities are modified for some selected pixels in this technique. It is not robust as it is prone to attacks by unauthorized users. In frequency domain technique [2], the image is transformed into the frequency domain and then watermark is embedded. Compared to LSB, it is more secured and robust to various attacks. In this approach, various transformation procedures are applied like Discrete Cosine Transformation (DCT) [3], Discrete Fourier Transformations (DFT) [4], Discrete Wavelet Transformation (DWT) [5], and Complex Wavelet Transformation (CWT) [6]. 2. PROPOSED METHOD There are three phases in this proposed scheme viz. Watermark generation, embedding, extraction and matching. Here the watermark is generated from the original image and the resolution of the watermark is on eighth of the original image. 2.1 Watermark generation The Watermark is generated by the following procedure:  Let the original image P is of size N x N  From P, randomly N-1 elements have been selected so that all the pixels have valid 3x3 neighborhoods.  Let the neighbors are: P(x-1,y-1),P(x-1,y),P(x-1,y+1), P(x,y+1), P(x+1,y+1), P(x+1,y), P(x+1,y-1), P(x,y-1)  Now mean value [say P a (x, y)] of those neighborhoods is calculated.  A binary sequence B i is obtained by applying the following condition: 0 if P(x, y) > P a (x, y) B i = 1 otherwise Where i = 1, 2, 3……N-1 A 8X8 Toeplitz matrix has been generated from the 8 bit binary sequence B i obtained above, which is used as the watermark to be embedded in the original image. Toeplitz Matrix: Toeplitz matrices [7] are matrices having constant entries along their diagonals. Toeplitz matrices arise in many different theoretical and applicative fields, in the mathematical modeling of all the problems where some sort of shift invariance occurs in terms of space or of time. The Toeplitz structure may occur entry-wise, for one-dimensional problems, or block-wise, for two-dimensional problems, or even at several nested levels in multidimensional problems. Toeplitz matrices may be finite or even infinite according to the features of the problem that is modeled. Given 2n-1 numbers a k , where k= -n+1,... -1, 0, 1, ..., n-1, a Toeplitz matrix is of the following form: